A model segmentation printing method based on self-supporting determination and transition zone division

Abstract

The application discloses a model segmentation printing method based on self-supporting determination and transition zone division, and belongs to the technical field of additive manufacturing. The self-supporting condition between printing layers of a model is determined by using voxels, and the overall model is preliminarily segmented into a self-supporting zone and a hanging flow zone by using clustering analysis on spatial positions. Considering that voxels in the interface between the hanging flow zone and the self-supporting zone are irregularly distributed, the self-supporting condition of the hanging flow zone cannot be determined by using an updated printing base surface, and a transition zone division method is given, wherein an adjacent voxel set obtained by using an octree and a printing base surface obtained by fitting are used to realize the division of the transition zone. By using the segmentation printing method, the model is segmented based on gravity effect, and the stability of additive manufacturing in the adjacent area of the self-supporting zone and the hanging flow zone is improved.

Description

Technical Field

[0001] This invention belongs to the field of additive manufacturing technology, and in particular relates to a model segmentation and printing method based on self-support determination and transition zone division. Background Technology

[0002] Additive manufacturing technology requires reducing or even avoiding the generation of support structures during the printing process. Traditional printing modes with fixed deposition directions have weak printing capabilities. For parts with complex structures, a large number of support structures need to be added to assist in the forming process, which can have a detrimental effect on the quality of the forming. In particular, in the field of metal additive manufacturing, supportless printing modes have become a key area of ​​research in printing technology.

[0003] For complex structural components, achieving unsupported layering is difficult. A segmentation approach can be considered, decomposing the complex solid into multiple sub-solids that can be formed in one step. Current methods, while using surface information to determine the flow area of ​​the model, struggle to further plan specific segmentation surfaces. Therefore, voxels can be used for self-support determination. During segmentation, the mainstream approach is to decompose the model only into self-supporting and flow areas. However, the printing quality at the boundary between these areas is generally poor, which also affects the final self-support determination result. Therefore, further defining partitions is necessary to achieve overall model segmentation and additive manufacturing path planning. Summary of the Invention

[0004] This invention provides a model segmentation and printing method based on self-support determination and transition zone division, solving the problem of flow sag in local areas during workpiece printing in traditional additive manufacturing processes. This invention uses a voxelized model for overall segmentation, employing a two-layer, voxel-by-voxel flow sag relationship determination to divide the overall self-supporting and flow sag regions. Simultaneously, it utilizes octree construction and fitting of the printing datum surface to divide the transition region. The entire model is segmented into multiple mutually separate self-supporting regions, flow sag regions, and transition regions.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A model segmentation and printing method based on self-support determination and transition zone division includes the following steps:

[0007] (1) Use regular hexahedral units to realize the voxelization of the STL model and set all voxel points to be segmented.

[0008] (2) Determine the printing layer to which each voxel belongs based on the printing base surface;

[0009] (3) Starting from the printing base surface, traverse each printing layer, determine the self-support status of the voxels, and divide the model into self-supporting areas and flow areas.

[0010] (4) Set all voxel points in the self-supporting region to the segmented state. If there are no voxel points in the hanging flow region, it means that the imported voxel set to be segmented has been segmented and step (8) is executed. Otherwise, if there is a hanging flow region, perform cluster analysis between voxels for the hanging flow voxels in the hanging flow region and divide the voxels into mutually separated sub-hanging flow regions according to the spatial position relationship of the voxels.

[0011] (5) Extract voxels that have face-to-face adjacency with the voxel set of the self-supporting region in each sub-flow region, and construct the adjacency voxel set of each sub-flow region;

[0012] (6) Based on the adjacent voxel set, the normal vector of the printing base surface of each sub-flow region is obtained by fitting using the least squares method;

[0013] (7) Determine the segmented transition area based on the printing base surface and adjacent voxel set, set the voxels in the transition area to the segmented state, remove the voxels in the transition area from the corresponding sub-flow area, and determine the position of the printing base surface corresponding to each sub-flow area in space.

[0014] (8) If there are no voxels in the state of being to be segmented, the final layering scheme is generated based on the segmentation results of the voxels and the corresponding printing base planes; otherwise, the remaining voxels to be segmented are clustered and step (2) is executed.

[0015] Furthermore, in step (2), the specific method for calculating the printing layer to which each voxel belongs is as follows: calculate the distance from each voxel to the printing base surface, and divide the printing layer according to the size of the distance.

[0016] Furthermore, in step (3), for the initial printed layer that is in direct contact with the printing base surface, it is assumed that all voxels in this layer are self-supporting; for the remaining printed layers, it is necessary to determine whether they are self-supporting with the adjacent printed layers below, specifically as follows:

[0017] Project the center points of all lower-layer voxels onto the printing base surface. Sequentially extract the center points of the current layer voxels and project them onto the printing base surface. Determine whether there is a projection point of the center of the lower-layer voxel within the r-neighborhood of the projection point of the current layer voxel. If there is, it indicates that the voxel belongs to the self-supporting voxel and is assigned to the self-supporting region; otherwise, it belongs to the floating voxel and is assigned to the floating region. Use a kd-tree to accelerate the search process of projection points in the neighborhood.

[0018] Furthermore, in step (4), the further division of the hanging flow area specifically includes:

[0019] Voxels are represented by voxel centroids. The DBSCAN algorithm is used to perform cluster analysis on the voxel centroid set. Based on the spatial relationship of the voxels, the voxels are divided into mutually separate sub-flow regions, thus completing the further segmentation of the flow regions.

[0020] Further, in step (5), the octree data structure is used to realize the search for adjacent voxels between the self-supporting region and the hanging flow region. First, the octree structure is used to store all voxels of the self-supporting region. Each leaf node of the octree corresponds to a voxel of the self-supporting region. Then, for each sub-hanging flow region, it is determined whether each hanging flow voxel is within the six-neighborhood of the leaf node of the self-supporting octree. If it is, it indicates that the hanging flow voxel is an adjacent voxel.

[0021] Further, in step (7), based on the obtained printing base surface normal vector, a printing base surface position is pre-defined, satisfying that all voxels in the flow region are on the same side of the printing base surface, and the average distance from adjacent voxel points to the printing base surface is less than or equal to the average distance from all flow region voxels to the printing base surface; the printing layer to which the adjacent voxels belong is calculated, and the printing layer to which all adjacent voxels belong is defined as the transition region, and the voxels in the transition region are set to the segmented state, while all voxels in the transition region need to be removed from the flow region; the position of the printing base surface is determined as the interface between the transition region and the flow region.

[0022] Furthermore, when determining the self-supporting relationship between the current layer voxel and the lower layer, the neighborhood radius r of the projection point is selected as the side length value of the regular hexahedral voxel.

[0023] Furthermore, in the DBSCAN algorithm, the neighborhood radius is set to 1.2 times the side length of a regular hexahedron voxel, and the minimum number of points is set to 6.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] This invention utilizes voxels to determine the self-support of the overall printed model. By employing neighborhood search at voxel center points, the overall model is efficiently divided into a self-supporting region and a flow region. Considering that when the model structure is relatively complex, voxels at the interface between the flow region and the self-supporting region may exhibit irregular distribution, making it impossible to perform further self-support determination on the flow region using the updated printing base surface, this invention proposes a definition and corresponding partitioning method for a transition region. This transition region is further segmented within the flow region using the clustering relationship of voxel center points and the adjacency relationship between voxels. The transition region ensures stable printing of the areas adjacent to the self-supporting region and the flow region, and allows for the continuous and stable determination of voxel self-support. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the model segmentation and printing method based on self-support determination and transition zone division according to an embodiment of the present invention.

[0027] Figure 2 This is a schematic diagram of the workpiece model selected for an embodiment of the present invention.

[0028] Figure 3 This is a schematic diagram of the clustering analysis results after the first self-support judgment in an embodiment of the present invention.

[0029] Figure 4 This is a schematic diagram of the final segmentation result in an embodiment of the present invention.

[0030] Figure 5 This is a schematic diagram of the final layering result in an embodiment of the present invention. Detailed Implementation

[0031] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the embodiments described below are intended to facilitate the understanding of the present invention and do not constitute any limitation thereof.

[0032] like Figure 1 As shown, a model segmentation and printing method based on self-support determination and transition zone division includes the following steps:

[0033] Step 1: Obtain the model file to be processed, and use regular hexahedral elements to voxelize the STL model, setting all voxel points to the state to be segmented.

[0034] In this embodiment, the model file to be processed is as follows: Figure 2 As shown. The voxel dimensions are 0.2mm * 0.2mm * 0.2mm.

[0035] Step 2: Determine the printing layer to which each voxel belongs based on the printing base surface.

[0036] In this embodiment, the printing layer thickness is set to 0.2 mm, and the distance from each voxel to the printing substrate is calculated to determine the printing layer to which each voxel belongs. For the initial voxel model, the printing substrate is the upper surface of the printing substrate.

[0037] Step 3: Starting from the printing base surface, traverse each printing layer, determine the self-support of the voxels, and divide the model into self-supporting areas and flow areas.

[0038] For the initial printed layer that is in direct contact with the printing substrate, all voxels in this layer can be considered self-supporting. For the remaining printed layers, it is necessary to determine whether they can be self-supporting with the adjacent printed layers below. This determination is mainly based on the gravity deposition effect. In the additive manufacturing process, the already formed part of the lower layer must be able to provide effective support for the part to be printed in the current printed layer to ensure the stability of the printing process. If effective bottom support cannot be provided, the part to be printed in the current printed layer will experience runoff, thus affecting the molding quality. The specific determination method is as follows:

[0039] Project the center points of all lower-layer voxels onto the printing base. Sequentially extract the center points of the current-layer voxels and project them onto the printing base. Determine if a projection point of the center point of the current-layer voxel exists within its r-neighborhood. If it does, the voxel is a self-supporting voxel and is assigned to the self-supporting region; otherwise, it is a floating voxel and is assigned to the floating region. A kd-tree is used to accelerate the search for projection points within the neighborhood.

[0040] In one implementation, the neighborhood radius r is generally determined by the size of the voxel and the self-supporting capacity of the material, and is typically selected as the side length of a regular hexahedral voxel. In this embodiment, r is 0.2 mm.

[0041] Step 4: Set all voxel points in the self-supporting regions to the segmented state. If there are no voxel points in the overflow region, it means that the imported voxel set to be segmented has been segmented, and proceed to Step 8. Otherwise, if there is an overflow region, perform cluster analysis between voxels for the overflow voxels in the overflow region, and segment the voxels into mutually separate sub-overflow regions according to the spatial positional relationship of the voxels.

[0042] In actual printing, if the voxels in the print run are not spatially continuous, the print run needs to be further divided to ensure the continuity of the printing path and the rational planning of the printing direction.

[0043] As one implementation method, the center point of a voxel represents the floating fluid element in the floating region, and cluster analysis is performed on the set of voxel center points to divide the voxel into spatially separated sub-floating regions.

[0044] In this embodiment, the DBSCAN algorithm is used, the neighborhood radius is set to 1.2 times the side length of the hexahedral voxel, and the minimum number of points is set to 6.

[0045] After the initial cluster analysis, the overall model can be divided into one self-supporting region and two sub-flow regions. The cluster analysis results after the initial self-support determination are as follows: Figure 3 As shown.

[0046] Step 5: Extract voxels in each sub-flow region that have face-to-face adjacency with the voxel set of the self-supporting region, and construct the adjacency voxel set of each sub-flow region.

[0047] In one implementation, an octree data storage structure is established for all voxel information in the self-supporting region, with each leaf node of the octree corresponding to a voxel in the self-supporting region. The octree is used to extract voxels in each sub-flow region that have face-to-face adjacency with the voxel set in the self-supporting region. Then, for each sub-flow region, it is determined whether each flow element is within the six-neighborhood of a leaf node of the self-supporting octree. If so, it indicates that the flow element is an adjacent voxel, thereby constructing the adjacent voxel set for each sub-flow region.

[0048] Step 6: Based on the adjacent voxel set, use the least squares method to fit the normal vector of the printing base surface of each segmented flow region.

[0049] Step 7: Determine the segmented transition region based on the printing base plane and adjacent voxel set, set the voxels in the transition region to the segmented state, remove the voxels in the transition region from the corresponding sub-flow regions, and determine the position of the printing base plane corresponding to each sub-flow region in space.

[0050] In one implementation, based on the obtained printing base plane normal vector, a pre-defined position for the printing base plane is established, ensuring that all voxels in the flow region are on the same side of the printing base plane, and the average distance from adjacent voxel points to the printing base plane is less than or equal to the average distance from all flow region voxels to the printing base plane. The printing layer to which adjacent voxels belong is calculated, and all printing layers to which adjacent voxels belong are defined as transition zones. Voxels in the transition zones are set to a segmented state, and all voxels in the transition zones need to be removed from the flow regions. The position of the printing base plane is determined as the interface between the transition zone and the flow region.

[0051] Step 8: If no voxels are in a state of being segmented, generate the final layering scheme based on the voxel segmentation results and the corresponding printing base planes. Otherwise, find the remaining voxels to be segmented and perform step 2.

[0052] The final segmentation result of this embodiment is as follows: Figure 4 As shown, the stratification results are as follows Figure 5 As shown.

[0053] It will be understood by those skilled in the art that the above descriptions are merely preferred examples of the invention and are not intended to limit the invention. Although the invention has been described in detail with reference to the foregoing examples, those skilled in the art can still modify the technical solutions described in the foregoing examples or make equivalent substitutions for some of the technical features. All modifications and equivalent substitutions made within the spirit and principles of the invention should be included within the scope of protection of the invention.

Claims

1. A model segmentation and printing method based on self-support determination and transition zone division, characterized in that, Includes the following steps: (1) Use regular hexahedral units to realize the voxelization of the STL model and set all voxel points to be segmented. (2) Determine the printing layer to which each voxel belongs based on the printing base surface; (3) Starting from the printing base surface, traverse each printing layer, determine the self-support status of the voxels, and divide the model into self-supporting areas and flow areas; for the initial printing layer that is in direct contact with the printing base surface, it is assumed that all voxels in this layer can be self-supported; for the remaining printing layers, it is necessary to determine whether they can be self-supported with the adjacent printing layers below, specifically: Project the center points of all lower-layer voxels onto the printing base surface. Sequentially extract the center points of the current layer voxels and project them onto the printing base surface. Determine whether there is a projection point of the center of the lower-layer voxel within the r-neighborhood of the projection point of the current layer voxel. If there is, it indicates that the voxel belongs to the self-supporting voxel and is classified into the self-supporting region; otherwise, it belongs to the floating voxel and is classified into the floating region. Use a kd-tree to accelerate the search process of the projection point in the neighborhood. (4) Set all voxel points in the self-supporting region to the segmented state; if there are no voxel points in the flow region, it means that the imported voxel set to be segmented has been segmented and step (8) is executed. Otherwise, if there is a flow region, perform cluster analysis between voxels for the flow region voxels and divide the voxels into mutually separated sub-flow regions according to the spatial position relationship of the voxels. (5) Extract voxels in each sub-flow region that have face-to-face adjacency with the voxel set of the self-supporting region, and construct the adjacency voxel set of each sub-flow region; (6) Based on the adjacent voxel set, the normal vector of the printing base surface of each sub-flow region is obtained by fitting using the least squares method; (7) Determine the segmented transition area based on the printing base surface and adjacent voxel set, set the voxels in the transition area to the segmented state, remove the voxels in the transition area from the corresponding sub-flow area, and determine the position of the printing base surface corresponding to each sub-flow area in space. In step (7), based on the obtained printing base surface normal vector, a printing base surface position is given in advance, satisfying that all the hanging flow voxels are on the same side of the printing base surface, and the average distance of adjacent voxel points to the printing base surface is less than or equal to the average distance of all hanging flow voxel points to the printing base surface. Calculate the printing layer to which adjacent voxels belong, define the printing layer to which all adjacent voxels belong as the transition region, set the voxels in the transition region to the segmented state, and remove all voxels in the transition region from the overflow region. The printing base surface is located at the interface between the transition zone and the drip zone; (8) If there are no voxels in the state to be segmented, the final layering scheme is generated based on the segmentation results of the voxels and the corresponding printing base planes. Otherwise, the remaining voxels to be segmented are clustered and step (2) is executed.

2. The model segmentation and printing method based on self-support determination and transition zone division as described in claim 1, characterized in that, In step (2), the specific calculation method for the printing layer to which each voxel belongs is as follows: calculate the distance from each voxel to the printing base surface, and divide the printing layer according to the size of the distance.

3. The model segmentation and printing method based on self-support determination and transition zone division as described in claim 1, characterized in that, In step (4), the further division of the hanging flow area specifically includes: Voxels are represented by voxel centroids. The DBSCAN algorithm is used to perform cluster analysis on the voxel centroid set. Based on the spatial relationship of the voxels, the voxels are divided into mutually separate sub-flow regions, thus completing the further segmentation of the flow regions.

4. The model segmentation and printing method based on self-support determination and transition zone division as described in claim 1, characterized in that, In step (5), the octree data structure is used to search for adjacent voxels in the self-supporting region and the hanging flow region. First, the octree structure is used to store all voxels in the self-supporting region. Each leaf node of the octree corresponds to a voxel in the self-supporting region. Then, for each hanging flow region, it is determined whether each hanging flow voxel is within the six-neighborhood of the leaf node of the self-supporting octree. If it is, it indicates that the hanging flow voxel is an adjacent voxel.

5. The model segmentation and printing method based on self-support determination and transition zone division as described in claim 1, characterized in that, When determining the self-supporting relationship between the current layer voxel and the lower layer, the neighborhood radius r of the projection point is selected as the side length value of the regular hexahedral voxel.

6. The model segmentation and printing method based on self-support determination and transition zone division as described in claim 3, characterized in that, In the DBSCAN algorithm, the neighborhood radius is set to 1.2 times the side length of a regular hexahedron voxel, and the minimum number of points is set to 6.

Images